The stakes are considerable. Weeds cost soybean growers more than 30 percent of their yield in severe cases, translating to an estimated 33 billion U.S. dollars in annual global losses. Worse, infestations often appear in scattered patches rather than uniform blankets, meaning a broadcast herbicide application treats vast stretches of weed-free ground. Depending on how weeds are distributed, spot spraying can cut herbicide use by as much as 66 percent when weeds cluster in small patches, though savings drop to roughly 10 to 20 percent when weeds spread evenly across a field. Existing ground robots can perform targeted spraying, but they struggle in wet soil, on slopes, or in small and irregularly shaped fields, and they must traverse large weed-free areas to reach isolated outbreaks.

The Arkansas team, led by Md Nurul Azmir and Cengiz Koparan, took a deliberately different route from the deep-learning mainstream. Instead of training a convolutional neural network to distinguish soybean from weed pixel by pixel, an approach that demands large annotated datasets and costly retraining for every new crop, growth stage, or field, their system quantifies total vegetation coverage in each image using a training-free segmentation method based on the Hue, Saturation, Value (HSV) color space. Pixels falling within calibrated HSV ranges are classified as vegetation, and coverage is simply the proportion of vegetation pixels in the image footprint. The clever twist lies in the reference: vegetation coverage measured from weed-free soybean plots at the same growth stage serves as a baseline, and any coverage exceeding that baseline is interpreted as a proxy for weed pressure.

The hardware is equally pragmatic. A MAPIR Survey3N multispectral camera, recording near-infrared, red, and green bands, feeds images to an NVIDIA Jetson Orin Nano, a compact edge-computing board with a 1024-core GPU and 8 GB of memory that draws at most 25 watts. When estimated weed pressure crosses a predefined threshold, the Jetson issues a pulse-width modulation (PWM) signal to an electronic speed controller, which drives a 12-volt brushless pump rated at 3.0 liters per minute and feeding two TeeJet nozzles. The entire sensing-to-actuation chain, from image capture through segmentation, decision logic, and pump activation, runs onboard, independent of the drone’s flight controls, with the DJI AGRAS MG-1S serving purely as the aerial carrier.

Field validation took place at the Milo J. Shult Agricultural Research and Extension Center near Fayetteville, where twenty soybean plots were planted in July 2024 with no herbicides applied, allowing natural weed populations to develop. Five plots were kept weed-free by hand to establish the crop reference. Across 1,000 images collected 14 and 24 days after planting, weed-free soybean coverage averaged 3.6 percent at the early date and 14.2 percent at the later one, but with substantial plot-to-plot and image-to-image variability, coefficients of variation reaching above 50 percent. That variability matters: a spray threshold set at the simple mean of weed-free coverage would trigger spraying on healthy crops, so the researchers recommend basing thresholds on the average of plot-level maximum coverage values, with a safety margin.

Threshold sensitivity analysis revealed a growth-stage-dependent tradeoff familiar to site-specific weed management. At 14 days after planting, when weed-free and weedy coverage distributions overlapped heavily, raising the threshold from 5 to 7.5 percent cut potential false spray activation from about 21 percent to essentially zero, but also collapsed the spray-trigger rate in weed-infested plots from 52 percent to 15 percent. By 24 days after planting, the distributions had separated enough that a 25 percent threshold produced only 4.8 percent false activation while still triggering sprays on 88.3 percent of weed-infested footprints, with a potential treatment-footprint reduction of 32.6 percent. In other words, the system’s discriminative power improves as the canopy develops, a finding that could let growers time their spot-spraying missions for maximum selectivity.

Indoor tests under controlled lighting confirmed the control logic end to end. Using green reference markers to simulate crops and weeds, the system correctly kept its pump off over bare background and crop-only configurations, then activated automatically when simulated weeds pushed total vegetation coverage to 14 percent and the weed-pressure proxy to 64.9 percent. Across 147 spray-actuation trials, the system matched the visually observed pump state 143 times, an accuracy of 97.3 percent, with precision of 99.1 percent and only three false negatives. System-estimated weed pressure tracked marker-based reference values with a coefficient of determination of 0.91, though a positive bias of 7.3 percentage points, traced to edge pixels around marker boundaries, suggests room for refinement in threshold calibration.

Field performance was more modest but still encouraging. Comparing system-derived vegetation coverage against manual ImageJ segmentation of the same images yielded a coefficient of determination of 0.69, a mean absolute error of 2.20 percentage points, and a slight negative bias of 2.02 percentage points. The gap between indoor and field agreement reflects the messy reality of real canopies: mixed crop-weed arrangements, variable illumination, shadows, and soil background all conspire against fixed color thresholds. During supplementary UAV flight tests, the sprayer sometimes failed to activate, likely because shadows and changing solar angle pushed computed coverage below threshold, or because motion blur from flight and rotor downwash degraded image quality. The authors are candid that these environmental factors remain the principal obstacle to reliable airborne operation.

The team is equally clear about what the study does not yet show. Spray deposition, droplet coverage, off-target drift, actual herbicide savings, and end-to-end sensing-to-actuation latency were not quantified, and pump-flow calibration alone does not characterize nozzle-level spray performance, which depends on operating pressure, droplet size, and effective spray width. Future validation plans call for water-sensitive papers placed at detected spray and no-spray locations to verify placement accuracy, alongside georeferenced detection outputs. Payload and battery constraints on small UAVs will also shape how much ground each mission can cover and how much spray volume it can carry.

Even with those caveats, the study makes a compelling case that vegetation coverage relative to a crop-only baseline is not merely a mapping metric but an actionable control variable, closing a loop that much weed-sensing research leaves open. By sidestepping species-level classification, the framework runs on modest hardware without labeled data, making it adaptable to new crops, growth stages, and geographies simply by recalibrating HSV thresholds and re-measuring a weed-free reference. The authors envision the approach extending to multi-platform systems in which drones scout rapidly and hand off targeted intervention to ground robots with greater payload and endurance. For an industry wrestling with herbicide resistance, environmental contamination, and the economics of treating weed-free ground, a drone that sprays only what needs spraying is a proposition whose time, and technology, may finally be converging.

Subject of Research: Adaptive UAV-based spot spraying for weed management in soybean using vegetation coverage thresholds

Article Title: Development of an adaptive sprayer control system for UAV-based spot spraying in soybean

Article References: Azmir, M. N., Tagoe, A., Runkle, B. R., Wang, D., Burgos, N. R., & Koparan, C. (2026). Development of an adaptive sprayer control system for UAV-based spot spraying in soybean. Smart Agricultural Technology, 15, Article 102589. https://doi.org/10.1016/j.atech.2026.102589

Image Credits: AI Generated

DOI: 10.1016/j.atech.2026.102589

Keywords: precision agriculture, UAV spraying, spot spraying, weed management, soybean, HSV segmentation, edge computing, variable-rate application, herbicide reduction, vegetation coverage, machine vision, sustainable farming

The central problem the researchers set out to address is a mismatch of scales and definitions. Laboratory porosimetry, and especially helium porosimetry on core plugs, provides accurate reference values for pore parameters, but these are discrete point measurements that do not adequately capture the spatial variability of pore fabrics across a reservoir. Digital image analysis, by contrast, offers a non-destructive way to quantify the visible pore space in thin-section images of rock. The catch is that image analysis yields a two-dimensional geometric porosity value that says nothing about whether the pores it counts are actually connected, accessible to fluids, or partially blocked by clay minerals that coat and choke pore throats. A rock can look porous in a photograph yet perform poorly as a reservoir.

The new workflow, which the authors call DIA–FIS, was tested on sandstone samples from the Petro-Masilah reservoir. The first stage relies on thin sections impregnated with blue epoxy, a standard petrographic preparation in which the resin fills the pore space and renders it a distinctive color under the microscope. The researchers applied pre-fixed thresholds in the hue–saturation–value color space, a representation of digital images that separates chromatic content from brightness and is often more robust to uneven illumination than working in red–green–blue channels alone. Segmentation in this space allows pore space to be isolated from grains and cements, and the resulting image-derived porosity can be calibrated against measured values.

A crucial step in the study was an explicit test of how sensitive this segmentation is to the choice of threshold boundaries. Because any threshold-based segmentation involves a judgment about where pore pixels end and grain pixels begin, the researchers conducted stability tests by perturbing their hue–saturation–value thresholds by plus or minus five percent and plus or minus ten percent. This kind of perturbation analysis quantifies the uncertainty that is usually left implicit when an analyst simply picks a threshold by eye. It also addresses a known weakness in the broader literature: image processing variability, including the choice between conventional thresholding and machine learning approaches, can meaningfully change pore-structural results.

The second stage of the workflow takes the calibrated image-derived porosity together with the measured clay volume ratio and feeds both into a Mamdani-type fuzzy inference system. Mamdani systems, one of the classic architectures of fuzzy logic, operate on linguistic variables and human-readable IF–THEN rules, in contrast to black-box machine learning models whose internal reasoning cannot easily be inspected. In this framework, statements such as rules linking higher image porosity and lower clay content to better reservoir quality are encoded explicitly, and the degree to which each rule fires is computed through membership functions that allow partial, gradual membership in categories such as low, medium, and high.

The output of the fuzzy stage is a defuzzified score the authors denote S_FIS, a dimensionless reservoir-quality score derived from the combined interpretation of image-derived porosity and clay volume. The authors are careful to define what this score is not: it is a rule-based assessment of reservoir quality rather than a physical porosity measurement, and it is not equivalent to effective, connected, non-clay, or helium-measured porosity. This distinction matters, because conflating a fuzzy quality index with a measured porosity would be a serious methodological error. The fuzzy output instead captures the gradual transitions between pore types and reservoir-quality classes, the very vagueness that hard classification schemes handle poorly.

The performance of the digital image analysis stage was evaluated against helium porosimetry performed on the same core plugs, one of the most reliable reference measurements available in petrophysics. The comparative analysis demonstrated good correlation for the image analysis stage, with a coefficient of determination greater than 0.95 and a mean absolute deviation below 3.5 percent. Those figures indicate that, once calibrated, simple threshold-based image segmentation tracks laboratory measurements closely for the visible pore fabric captured in two dimensions, even before any fuzzy reasoning is applied.

The fuzzy stage then did what it was designed to do: it penalized clay. Samples classified as clay-affected received lower reservoir-quality scores and lower quality classes from the inference system, while cleaner sandstones received more favorable interpretations. This behavior reflects genuine petrophysical reasoning. Clays, whether dispersed between grains or lining pore walls, reduce effective pore throats, impair permeability, and make a rock’s total porosity a misleading indicator of its fluid-storage and fluid-delivery capacity. By explicitly including clay volume as an input to the rule base, the model forces the interpretation to account for a factor that a purely geometric porosity number would ignore.

Interpretability is the distinguishing selling point of the approach. Deep learning methods have made remarkable inroads into digital rock physics, from segmenting digital rock images to predicting porosity from well logs, but their opacity can be a barrier in an industry where decisions worth millions ride on petrophysical interpretations and where regulators and partners may demand to know why a model produced a given answer. A Mamdani fuzzy system, by contrast, can be audited rule by rule: a geologist can read the IF–THEN statements, check whether they accord with established reservoir engineering understanding, and adjust membership functions if local rock properties demand it. The workflow thus positions itself as a transparent complement to, rather than a replacement for, conventional laboratory characterization.

The authors describe the DIA–FIS workflow as a reproducible, non-destructive complement to traditional reservoir characterization, and the framing is apt. The image analysis can be repeated on archived thin sections without consuming additional core material, the threshold perturbation tests make the segmentation uncertainty explicit rather than hidden, and the fuzzy rules encode petrophysical reasoning in a form that domain experts can verify and refine. For heterogeneous, clay-influenced sandstone reservoirs such as those of the Sayun-Masilah region from which the study samples derive, the approach offers a way to extract more interpretable information from routine petrographic preparations. The dataset underpinning the study comprises petrographic and chemical characterization of sandstone samples from the Petro-Masilah reservoir; because the core material is proprietary, the data are curated by the corresponding author and available upon reasonable request for academic and non-commercial research purposes. As machine learning continues to spread through the geosciences, studies like this one make the case that sometimes the most valuable model is not the most powerful one, but the one whose reasoning a reservoir engineer can actually read.

Subject of Research: Hybrid digital image analysis and fuzzy logic for porosity and reservoir-quality estimation in clay-influenced sandstone reservoirs

Article Title: Interpretable hybrid digital image analysis model with fuzzy logic system for porosity estimation in clay-influenced sandstone reservoirs

Article References: Interpretable hybrid digital image analysis model with fuzzy logic system for porosity estimation in clay-influenced sandstone reservoirs. (n.d.). https://doi.org/10.1007/s12145-026-02230-7

Image Credits: AI Generated

DOI: 10.1007/s12145-026-02230-7

Keywords: porosity estimation, digital image analysis, fuzzy logic, fuzzy inference system, sandstone reservoirs, thin sections, HSV segmentation, clay content, reservoir quality, helium porosimetry, petrophysics, interpretability